Do you even compress bro?
Many people tend to avoid the topic of compression.
They think it’s complicated and ultimately not worth it.
This is especially true for beginners.
In reality, it is not so much about the degree of difficulty though.
Understanding the relationship of the 4 core elements of every compressor is key:
Threshold, Ratio, Attack & Release.
Imagine your typical compressor like this:
No matter what plugin you use, every compressor consists of these four main components.
Some may have additional options and features but as long as you get acquainted with the “Fab 4”,
you will be on your way to compression mastery.
So what do they do in particular?
“Isn’t too happy about his job but knows it’s important: Mr. Threshold”
A threshold is a powerful tool.
It allows us to set a limit at which every frequency that crosses the line will get compressed.
Think of it like this:
“A line that should(not) be crossed: The threshold.”
The frequencies that stay below the threshold will remain unaffected.
The frequencies that cross the threshold, will get attenuated.
Or in other words: We divide the spectrum into two halves:
The upper half is the place in which everything gets compressed.
The lower half is the place in which everything remains unaffected.
By how much is everything that crosses the threshold compressed?
Good question! This is the only piece of information that our threshold doesn’t know about.
Luckily, Mrs. Ratio can help us out here.
“Finding the right ratio can be overwhelming.”
So far we know how to set the point at which the compression is supposed to take place.
The ratio, on the other hand, allows us to set the degree at which the frequencies will get compressed.
Or in other words. How “hard” the compression will be.
It’s great that we know WHERE the frequencies get compressed
but we also want to know by HOW MUCH.
Here’s an explanation:
If you set a ratio of 2:1 that means for every 2db that crosses the threshold, only 1db will actually come through.
Hence, the other 1db will get compressed.
“Happy that she made it – Decibel 1 out of 2”
As you can imagine, the higher we set the ratio, the harder the compression will be.
Many people dislike the ratio concept. The reason:
It is more difficult to visualize the amount of compression if it’s expressed as 2:1 or 6:1.
Do you want to know the PRECISE amount of compression that takes place?
Take a look at this:
It is very easy to break down ratios into percentages.
If you want to get a better idea of any ratio just follow this formula:
Desired Ratio: 2:1
Step 1#: Flip the numbers | 1:2
Step 2#: Divide the numbers | 0.5
Step 3#: Multiply by 100 | 50
Step 4#: Add a % | 50%
As you can tell, a 2:1 ratio is consequently a compression level of 50%.
That means the amount that crosses the threshold will be attenuated by 50%.
Or yet another way of looking at it:
2db make it across the threshold? Only 1db will survive.
4db make it across the threshold? Only 2db will survive.
10db make it across the threshold? Only 5db will survive.
It’s like a bad horror movie where you already know what’s about to happen…
Hence, a ratio of 4:1 equals a compression level of 25%
I created a graphic which includes all of the major ratios expressed in percentages for you:
Any ratio higher than 10:1 is basically a “Limiter”.
A Limiter compresses EVERYTHING that crosses the threshold.
If you look at the graph above you can see that a ratio of 10:1 already compresses 90% of the signal.
If you want a natural compression, you should stay below that ratio.
As a side note.
If we take things very literal, a 10:1 ratio still wouldn’t be considered “Limiting” by some people, because of the fact that there’s still 10% of the signal coming through. In universities, you often learn that a ratio of 20:1 or 25:1 is considered Limiting.
While a ratio of 20:1 only allows 5% of the signal to pass through, we are nitpicking at this point.
Whether you take a ratio of 10:1 (10% of the signal passes) or 20/1 (5% of the signal passes), in both cases you have an extreme amount of compression that would be considered as Limiting by most people. Ultimately it is up to you to decide when you feel the compression is Limiting.
Before we move on to the children of Mr. Threshold and Mrs. Ratio, here’s a final graphic which visualizes the compression a little bit further:
At this point, we know how to set the threshold and the ratio.
Or in other words: WHERE signals get compressed and by HOW MUCH.
The only question left is WHEN?!
When can I haz frequencies?
The attack or “attack time” defines the delay until the compression kicks in.
That is AFTER a signal has crossed the threshold!
If the attack knob could speak, it would only ask one question:
WHEN do I get to start the compression?
If you set a fast attack time, it means that the compression will kick in quickly.
If you set a slow attack time, the compressor will wait a pre-defined amount of ms (milliseconds) until it gets triggered.
Does it even matter to add a delay to the compression?
Absolutely! The attack allows us to target the transients and sustains of a signal very precisely.
Think of your audio signal as an “animal”.
Every audio signal has a “head” and a “tail”.
The head is the initial part of your audio.
The tail is the sustain.
Depending on your material, you may want to enhance or compress the head of your audio.
Let’s assume you have a piano in a mix.
If you’d want the piano to have more punch,
you could precisely target the audio by utilizing the attack knob.
Do you want the initial thumb, the transient, to stand out more?
Add a slow attack time and compress the tail.
That way the transients come through and only the sustain gets compressed.
After adding some makeup gain, your piano will have more punch.
If you only turn up the gain, you are increasing the total volume of the instrument.
This is however not always the best choice since you’d also increase the volume at parts that you don’t want.
The first waveform was increased in gain.
The second waveform shows the original, uncompressed file.
As you can see, it is not only the peaks that rise but also the quieter parts.
That’s why compression is such a useful tool!
If you feel that your piano sounds to aggressive, you could do the opposite:
Add a fast attack and compress the head.
That way the head of the piano gets tamed and the sustain will come through better without turning down the volume.
The attack allows us to determine precisely which part of the audio we want to compress.
“Let it go”
The release is basically the opposite of the attack:
While the attack determines when the audio is supposed to get compressed,
the release determines when the compression is supposed to let go.
Using the attack and release together gives you full control over the compression process.
Setting a fast release means that the compression will only last shortly.
The consequence is that the signal will be left uncompressed again after the release.
The slower you set the release, the longer the compression will be active.
A very slow release will result in a compressed signal EVEN when the peaks are below the threshold again.
The compression will essentially “jump” onto the peaks that cross the threshold
and not let go until the release time is over.
While you can easily set a fast attack, it is usually not recommended to set an extremely fast release.
A fast release can introduce unwanted artifacts. This is especially true if you set your release time close to the attack time.
In that case, the compressor wouldn’t really have a lot of time to react properly. After all, we are dealing with miliseconds.
Always be a bit more careful with your release time.